Several lines of research demonstrate that the immune system maintains a homeostatic balance of regulatory and effector activity. Loss or inhibition of the activity of regulatory T cells, by deficiency in IL-2, CD25, FoxP3, or aire, or b neonatal thymectomy, may result in widespread catastrophic autoimmunity mediated by T cells whose targets include tissues of the eye. Many tissue-specific antigens enjoy ectopic expression in the thymus, in part due to aire activity, providing the opportunity to generate a diverse repertoire of regulatory T cells to self-antigens, and protection from autoimmunity. These natural regulatory T cells (nTregs) are required for survival of the host. Tregs are also induced in the periphery from mature T cells (iTregs) upon encounter with antigens, both self and foreign, under conditions that promote Treg differentiation, including the presence of TGF and retinoic acid. In many cases these Treg-generating interactions take place in lymphoid tissues, including lymph nodes. There is also evidence that some tissues are populated by apparently resident Tregs; these tissues include skin, adipose tissue, and the gastro-intestinal tract where their local activity appears to contribute to tissue homeostasis. Our preliminary studies suggest that local generation of Tregs may be found during responses to retinal antigens, a process we refer to as on-demand Tregs. We have found a small population of T cells in the parenchyma of the normal retina. A subset of these T cells expresses a transgenic marker for Tregs using the foxp3 promoter. We propose that these Treg cells are generated locally, and are evidence of an on-going contribution to maintenance of immune homeostasis in the retina. We further propose that they are induced and maintained by the presence of local retinal dendritic cells with regulatory antigen presenting cell activity. These hypotheses are explored in the following Aims. Aim 1 asks if the presence of Tregs in the normal, quiescent retina is promoted by their expression of a T cell receptor with specificity for a retinal antigen. Aim 2 examines the hypothesis that the function of retinal Tregs is to protect the retina from EAU or other local inflammation and examines their phenotype before and after local injury or retinal inflammation. Aim 3 will test their origin, and examine our hypothesis that they are made on-demand.